OCR Text |
Show The alternative to the experimental technique is to model the combustion performance by a computer program. The main advantages of the computer modelling technique are the lower cost in the long run in comparison with experimental technique and less time-consuming. One drawback however wi th mode 11 i ng techn i ques is that the output of the program is only as good as the computer program itself. Therefore, there is strong requirement to ensure that the model gives a true representative of the combust i on process. At present a number of commerc i a 1 computational fluid dynamics (CFD) program codes are available and used extens i ve 1 y, such as FLUENT, TEACH, PHOEN IX, FLOW 3D etc. A 1 though all these CFD codes are capable of predicting the combustion performance such as temperature and vel oc i ty fi e 1 d, but they do not include a deta i 1 ed pollution emission model. Thus there is a strong requirement for a postprocessing computer· model which is capable of predicting pollutant emissions during combustion processes. This study reports the application rationale of such a computer code capable of predicting the NO emission during combustion of fossil fuels. There are a number of advantages in using post-processing models. Parallel development of modules decreases the program development and debugging time. The program can be easily modified by substituting more efficient routines or improved kinetics when they become available. The post-processing model can be made inactive while solving particular flow problems and therefore reducing the computing time. The developed code is based on the calculated temperature and velocity fields obtained from CFD code and predicts thermal, prompt and fuel-NO emissions during combustion processes. These cal cul at ions take into account the effect of turbulence/chemistry interaction on the rate of formation of NO. Finally the developed model is used for the prediction of the conditions inside an industrial regenerative burner used in steel industry. This study di scusses the compari son between experimental and theoret i cal NO 1 eve 1 s and in addition examines the influence of the burner design and xoxidant concentration on the level of NO emission from the burner. 2. MODELLING OF NO FORMATION IN STATIONARY COMBUSTORS Comprehensive and detailed chemical kinetic models for predicting NO emi ss ions duri ng combust i on processes have been under development since the 1 ate 1970' s as summari sed in more recent publ ications (1, 2). In their present form, the detailed models consist of typically 250 reactions involving 40-50 chemical species with the kinetic data being updated in accordance with recent advanced experimental measurements. Although these models have been available for some time, they continue to be used for advanced research purposes and have not recei ved general acceptance as tools for predicting NO emissions in industry. Some of the major reasons for reluctance to use these detailed kinetic modelling codes in industry for design applications are as follows: (a) an inability to use the comp 1 ex deta i 1 ed chemi ca 1 . k i net i cs for des i gn app 1 i cat ions as they are dif~icult to use in a design environment, (b) even if the objection (a) is resolved, using a detailed chemical mechanism for NO emission predictions will require a powerful computer and significant amount of computer time wh i ch wi 1 1 make use of these models economi cally un attract i ve and (c) a scepticism on the part of design engineers to accept the present models to the point that they can predict accurately the complex processes of NO emission which is controlled not only by a chemical mechanism but also the knowledge of turbulence effects on NO formation rate: 2 |